We need to preload the ancestors of search results after applying
pagination limits. This way the search results itself are paginated,
but not the ancestors.

If we don't do this, we might not preload a parent group of a search
result as it has been cut off by pagination.

* Group filtering now includes also issues/MRs from
subgroups/subprojects
* fixed due_date
* Also DRYed todo controller specs

Enum in Rails 5 does not map nil values - IOW nil value remains nil,
even if there is a key with nil value in the enum definition.

This commit overrides the underlying Enum methods so nil value is
still mapped. This solution is far from being ideal: it uses dynamic
definition of methods which introduces more magic/confusion
into the codebase.

It would be better to get rid of the nil value in enums.

This is tricky: the query was being run in
`ObjectStorage::Extension::RecordsUploads#retrieve_from_store!`, but we can't
just add batch loading there, because the `#upload=` method there would use the
result immediately, making the batch only have one item.

Instead, we can pre-emptively add an item to the batch whenever an avatarable
object is initialized, and then reuse that batch item in
`#retrieve_from_store!`. However, this also has problems:

1. There is a lot of logic in `Avatarable#retrieve_upload_from_batch`.
2. Some of that logic constructs a 'fake' model for the batch key. This should
   be fine, because of ActiveRecord's override of `#==`, but it relies on that
   staying the same.

Signed-off-by: NRémy Coutable <remy@rymai.me>

Caching the attributes as JSON and manually instantiating the record
ended-up very complex since the edge-cases such as upload fields,
serialized fields, and fields with custom accessors had to be handled.

To ensure 3 points out of 4 are checked from https://gitlab.com/gitlab-org/gitlab-ce/issues/45175
we now include the Rails version in the cache key.
Signed-off-by: NRémy Coutable <remy@rymai.me>

Fixes that make this work:

* A change in Ruby (https://github.com/ruby/ruby/commit/ce635262f53b760284d56bb1027baebaaec175d1)
requires passing in the exact required length for OpenSSL keys and IVs.

* Ensure the secrets.yml is generated before any prepended modules are
loaded. This is done by renaming the `secret_token.rb` initializer to
`01_secret_token.rb`, which is a bit ugly but involves the least impact on
other files.

When deleting associated records, Rails loads all associations into memory
(https://github.com/rails/rails/issues/22510) before destroying them. This
can cause a surge in memory and cause destruction of objects to fail
due to idle in transaction database timeouts. This fix is inspired from
https://github.com/thisismydesign to destroy `has_many` relationships
in batches.

Closes #44610

We request Gitaly in a N+1 manner to build discussion diffs. Once the diffs are from different revisions, it's hard to make a single request to the service in order to build the whole response.
With this change we solve this problem and simplify a lot fetching this piece of info.

This can occur when a note is added to a commit and then this commit
is deleted.

Signed-off-by: NRémy Coutable <remy@rymai.me>

This reverts commit 2ebafdfb

This reverts commit 4e1bb1d1.

destroy_all loads all records at once

ObjectStore uploader requires presence of associated `uploads` record
when deleting the upload file (through the carrierwave's after_commit
hook) because we keep info whether file is LOCAL or REMOTE in `upload`
object.

For this reason we can not destroy uploads as "dependent: :destroy" hook
because these would be deleted too soon. Instead we rely on
carrierwave's hook to destroy `uploads` in after_commit hook.

But in before_destroy hook we still have to delete not-mounted uploads
(which don't use carrierwave's destroy hook). This has to be done in
before_Destroy instead of after_commit because `FileUpload` requires
existence of model's object on destroy action.

This is not ideal state of things, in a next step we should investigate
how to unify model dependencies so we can use same workflow for all
uploads.

Related to #45425

In Rails 5.0 automatic inverse does not work for polymorphic relathionships.
It was fixed in Rails 5.2: https://github.com/rails/rails/pull/28808

Until that the `inverse_of: :source` argument should be set explicitly.